On 6 August, the largest rover ever sent to Mars should touchdown on the planet’s surface. For the landing, the mission uses a complex – and untried – “sky crane”. Our space columnist finds out how it’s going to work. (Video of landing courtesy of Nasa).

If you’re contemplating a mission to Mars, Nasa’s Chronology of Mars Exploration makes depressing reading. For every successful Russian Mars 3 or American Viking 1, there’s a failed Beagle 2 (UK) or Mars Polar Lander (US). In fact, of the 42 missions listed, only 17 have been successful. The odds, over the past half-century, of a Mars mission succeeding are around 40%.

To be fair, several early missions didn’t even make it off the launch pad and the chances of success have improved considerably over the decades. But you don’t have to go back far to see failure. Only last year, Russia’s Phobos-Grunt mission to the Martian moon Phobos failed to make it out of Earth orbit (well, it did eventually when it burned up on re-entry). Perhaps the most infamous though is Nasa’s 1999 Mars Climate Orbiter, where a mix-up between imperial and metric measurements sent the spacecraft careering into the Martian atmosphere to be destroyed.

On 6 August (GMT) 2012 Nasa will try again with what is almost certainly the most ambitious and exciting Mars mission ever launched. The Mars Science Laboratory (MSL) with the Curiosity rover is designed to investigate whether the planet ever had the conditions to support life.

The rover is essentially a robotic geologist, that will collect and analyse soil and rock samples as it trundles across the Martian surface. And it is big: around the size of a Mini Cooper or small SUV; at 900kg (almost a tonne), it’s heavy and when hurtling towards the planet, it’s travelling at some 5km/s (mps), roughly 18,000km/h (11,000mph).

So, here’s the engineering challenge: successfully land a rapidly moving, car-sized rover on an alien planet. Remember to show your workings.

“When you have a big vehicle, it’s actually very difficult to slow down,” says Dan Rasky from Nasa Ames in Silicon Valley, California. Rasky, now the director of the Emerging Commercial Space Office, helped develop the MSL heat shield that will protect the spacecraft as it enters the Martian atmosphere.

“The major part of slowing down is being done with your heat shield,” says Rasky. “It’s a very tenuous atmosphere – similar to 100,000 feet [30,000 metres] here on Earth – but if you design things right, you can still slow things down enough that you can land safely.”

‘Scary’ idea

Nasa engineers originally planned to employ the same material that was used to land the Viking missions in 1975. But when they tested it in a special wind tunnel, equipped with high intensity heaters designed to simulate the conditions the MSL spacecraft will face, things didn’t turn out as expected.

“The material didn’t work,” Rasky tells me. “The very high heat just burned through the heat shield which would have burned into the structure of the spacecraft. So you would have got to the surface in pieces.”

Instead, the engineers turned to Phenolic Impregnated Carbon Ablator, or PICA for short. Phenolic is the same material that we use for saucepan handles, a plastic that burns but doesn’t melt. PICA was first used on Nasa’s Stardust mission and is also fitted to the SpaceX Dragon Capsule.

“There’s part of it that gets burned away. The Phenolic burns to generate a pyrolysis gas, which turns out to be an important way that it absorbs the heat,” Rasky explains. “SpaceX intends to get multiple uses out of its heat shield, which you can do if you size it correctly.”

Once the heat shield has done its job, the MSL spacecraft should have slowed to around 400m/s. With the rover still encased in the shell of the heat shield, the parachute deploys to slow the descent down even further. Parachutes have a proven track record on Mars – most recently with the Phoenix mission. But with MSL, the parachute is only the first stage in a much more complex landing process. This will be the first mission to use a ‘sky crane’.

“It looks scary but it actually eliminates a lot of the problems that previous designs had,” says Nasa’s Steve Sell, an engineer at the Jet Propulsion Laboratory in Pasadena who helped develop the concept.

On the face of it, the sky crane looks unnecessarily complicated. Once released from its parachute and protective shell, the lander uses engines to slow the descent. At this stage, the MSL resembles a flying bedstead with the rover slung underneath. As this combination nears touchdown, the rover is lowered towards the surface on cables and its wheels unfurl. On contact, the cables are severed and the skycrane (bedstead) flies away and crashes and the rover starts work.

What could possibly go wrong?

Silent entry

I ask Sell how they ended up with something that’s so complex. “When you’re designing rover style missions in particular, you quickly run into a problem where if you’re going to put the rover on top of a lander, with legs and rockets, you have to get it off there somehow when you’re going to touch down,” he explains.

“There’s a very complicated issue of designing a way to reliably, on all kinds of different circumstances – on slopes, on rocks, on sand dunes - drive something off the top of a platform. You also have the additional complexity that for Curiosity, this rover is over five times larger than any previous rover built to land on Mars.”

That means that if they put the rover on top of the lander, there’s every chance it would make the whole arrangement too top heavy, causing it to topple over. Likewise, if you attached the rover beneath the lander, it could get trapped if the legs sank into sandy ground or landed at an odd angle. So that’s out too.

Another option that’s been used successfully – most notably on the recent Spirit and Opportunity missions – is airbags. With this design, the rover bounces across the surface encased in a cocoon. When it comes to a stop, the airbags deflate, the cocoon unfolds and the rover trundles away. So why not use that system?

“When you start trying to make that [airbag system] bigger, so it could safely land a one tonne rover, you end up with an airbag system that is very, very large…it just doesn’t scale up.”

So, it was back to the drawing board. Eventually, by a process of elimination, they ended up with the sky crane.

“You can think of the rover wearing a jet pack on its back,” says Sell. “When we get 20m above the surface, the jet pack lowers the rover on three bridles and that whole system keeps moving down towards the surface with the rover held seven and a half metres below the jetpack. When it touches down, the jet pack stops, it cuts the bridles free, then the jetpack flies away and crashes...it’s a way of landing without ever having to land the spacecraft.”

The landing sequence from entry into the atmosphere, to the rover landing takes place automatically in just seven minutes. The last command sent to MSL will be two hours before touchdown and, because of the time delay between Earth and Mars, mission control won’t know whether it’s succeeded until some 14 minutes after.

“By the time we get the signal that says ‘I’ve started entry’ – it will already have been on the surface for seven minutes!”

So does all this make Sell nervous? “It actually makes me very confident. Most of us on the project have been involved in the design and testing of the spacecraft for years now…we understand how everything on the spacecraft was built and tested and put together.”

Rasky, though, is a little more cautious: “There’s always a risk that something won’t go well but we are certainly hoping for the best…there’s always that lingering doubt.”

If you would like to comment on this story or anything else you have seen on Future, head over to our Facebook page or message us on Twitter.

You can hear from a scientist working on the mission and more about the landing system in Richard’s latest podcast.

Around the BBC